Physical and Biological Barriers to Viral
Vector–mediated Delivery of Genes
to the Airway Epithelium
Raymond J. Pickles
Cystic Fibrosis/Pulmonary Research and Treatment Center, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
A gene therapy for cystic fibrosis (CF) lung disease by intralumenal delivery of therapeutic transgenes into the lung is a logical treat-ment strategy if efficient gene transfer can be achieved without detrimental effects to the patient. Indeed, pioneering work in the late 1980s showed that genetically engineered viruses could deliver the CF corrective transgene to cultured cells from patients with CF. However, after many attempts to deliver the corrective gene to the lungs of patients with CFin vivoand with the luxury of 20/20 hind-sight, it is realized that although logical, the strategy to accomplish this task did not appreciate the evolution of the lung to resist invasion by pathogens such as viruses. It is now apparent that several levels of barriers exist that restrict exogenous gene delivery to the airway epithelium by commonly used viral vectors. Components of the innate and cell-mediated immune system collectively limit both the access to and duration of gene transfer vectors to the airway epithelium. Alternative viral vectors that have evolved to circumvent these barriers will require further development if gene transfer is ever to be considered a therapy for CF lung disease.
Keywords:cystic fibrosis; gene therapy; viral vectors
In this short article, I will sum up experiences of my laboratory using adenoviral vectors (AdV) as gene delivery vehicles for the respiratory epithelium, the target tissue for the treatment of the underlying cause of cystic fibrosis (CF) lung disease. This work has mainly focused on the “innate” physical and biological barri-ers posed by the airway epithelium that limit gene transfer effi-ciency, with the opinion that until efficient gene transfer to the correct target cells can be achieved the prospect for a safe CF gene transfer strategy will be limited. This review will discuss mainly my conclusions from work performed in my laboratory in the context of work performed by others in the field. Additional barriers to effective and safe gene transfer such as the host cellular immune response against vectors and/or transgene anti-gens and the mechanisms of delivering gene transfer vectors to the lung will not be discussed here, and the reader is referred to several recent reviews that address these issues (1, 2).
Several separate scientific advances culminated in the idea that a gene transfer approach for CF lung disease would be feasible. Many years of basic research with a common lung virus, the adenovirus, had generated less pathogenic replication-deficient vectors capable of expressing a transgene of choice. Secondly, in 1989, the defective gene that results in CF disease was cloned and the normal product of this gene identified as a cAMP-activated chloride ion channel, the cystic fibrosis transmembrane conductance regulator (CFTR) (3).
(Received in original form March 10, 2004; accepted in final form May 25, 2004)
Correspondence and requests for reprints should be addressed to Raymond J. Pickles, Ph.D., Cystic Fibrosis/Pulmonary Research and Treatment Center, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7248. E-mail: branston@ med.unc.edu
Proc Am Thorac Soc Vol 1. pp 302–308, 2004 DOI: 10.1513/pats.200403-024MS
Internet address: www.atsjournals.org
The first signs of CF lung disease occur in the distal bronchio-lar airways. With time, bronchiolitis and mucus plugging of the small airways are common findings in the CF lung. The link between CF pathogenesis and the CFTR defect is related to the normal function of CFTR maintaining hydration of the pericili-ary fluid layer that lines the airway surface (4). An absence or expression of defective CFTR in the lung results in reduced chlo-ride ion secretion, hyperabsorption of sodium ions, increased vis-cosity of airway secretions, impaired mucociliary clearance, chronic bacterial infection, bronchiectasis, and premature death (5). Be-cause these pulmonary manifestations are likely primary or sec-ondary to loss of CFTR function, the most efficacious strategy to treat CF lung disease would be to replace normal CFTR function in airway epithelial cells, thus “correcting” lung epithe-lium homeostasis and hopefully lung function.
The airway epithelial cell types that exhibit all of the ion-and fluid-transporting functions of CFTR ion-and display abnormal function in patients with CF are the ciliated epithelial cells (6), which are considered to be the target cell types that require correction (7, 8). However, immunolocalization studies have identified the serous cells of the submucosal glands as the highest CFTR-expressing cell type in the lung, suggesting that these cells may also be an important target for gene replacement (9).
Shortly after the identification of the CFTR gene, two groundbreaking observations made gene therapy for CF lung disease appear imminent. First, isolated epithelial cells derived from the airway epithelium of patients with CF and cultured on plastic dishes could be phenotypically “corrected” by transfer-ring the CFTR cDNA into the cells (10). Secondly, replication-defective AdV engineered to express the CFTR cDNA were administered to the airways of experimental animals, and trans-gene expression was observed in respiratory epithelium (11). These pioneering studies produced a flurry of scientific activity and excitement in both the gene therapy and CF scientific com-munities, and within 3 years of these initial observations the first clinical trial describing AdV-mediated gene transfer to the airway epithelium of patients with CFin vivowas reported (12). Over a decade later, these promising early observations have unfortunately not withstood further investigation. After over 20 gene therapy clinical trials for CF lung disease (of which greater than 70% used AdV), the gene therapy community has realized that gene transfer to airway epitheliumin vivo, although logical, is not trivial. The evolution of the respiratory epithelium as an effective barrier to invading pathogens entering the lung (e.g., viruses), by a host of innate and cell-mediated immune systems, culminates in restricted cellular uptake of gene transfer vectors and reduced transgene expression.
VIRAL VECTORS AS GENE DELIVERY VEHICLES TO THE LUNG: AMBITIOUS OR FAR-SIGHTED?
of CF lung disease was first tested by Johnson and colleagues, who showed in CF cell–mixing experiments that 6 to 10% of CFTR-expressing cells were required to restore normal levels of chloride secretory function to an epithelium in vitro (13). However, this degree of “correction” was insufficient to correct the hyperabsorption of sodium ions that is likely to be necessary for resolving CF lung disease. To restore the normal sodium transporting capabilities of the respiratory epithelium, it has been hypothesized (but not proven) that greater than 80% of the surface epithelial cells will have to express CFTR (7). There-fore, the number of CF cells to be targeted by a vector, i.e., ciliated cells, required to express CFTR will be high. Even so, it must be emphasized that correction of the chloride secretory defect by replacement of functional CFTR has not yet been shown to restore ionic homeostasis to the CF lung or reverse CF-related pathology.
The safety of gene transfer to airway epitheliumin vivousing viral and nonviral vectors in clinical trials has so far been promis-ing. However, determination of the efficiencies of gene transfer in the trials performed to date have shown, at best, only partial “correction” (⬍20%) of the CF bioelectrical defect (12, 14–16). No attempts to determine efficacy of gene transfer on CF lung disease-related end-points have yet been attempted.
IDENTIFYING AND APPRECIATING THE TARGET CELL TYPES FOR CFTR DELIVERYIN VIVO
Adenoviral vectors are extremely efficient gene delivery vehicles once the virus has entered into the target cell. Why then is this respiratory virus apparently so inefficient at targeting the intact epitheliumin vivo? One consideration when comparing wild-type adenovirus infection to transduction by AdV is that the replication-deficient vectors rely on delivering many virus particles to a target tissue, whereas wild-type virus only requires access to a small number of cells from which it can propagate and spread within the target tissue. Replication-competent wild-type Ad may take advantage of regional differences in airway epithelium integ-rity and injured epithelium has been shown to be more susceptible to AdV transduction than intact epithelium (17).
Another misconception about AdV infection of cells was the correlation of gene transfer efficiency to isolated airway epithelial cellsin vitroto that expectedin vivo. The failure of the lung gene transfer community to appreciate the complex phenotypic and morphologic characteristics of the target airway epithelial cellsin vivoresulted in trying to “force-feed” viruses into the target cells without sufficient attempts to understand the mechanisms that were restricting viral entry into the cell.
Airway epithelial cells are present throughout the conducting airways of the lung, including the nasal, tracheal, bronchial, and bronchiolar regions. Airway epithelial cell type composition is complex because it is species- and airway region–dependent, and the reader is referred to comprehensive reviews describing species-specific airway epithelial cell distribution (18, 19). For the human lung, the conducting airways are composed of several epithelial cell types dependent on airway region. Generally, the surface epithelium is composed of ciliated cells, mucus-secreting cells (goblet), serous cells, Clara cells, and basal cells. Alveolar regions of the lung consist primarily of alveolar Type I and Type II cells, although these cell types are not thought to participate in the pathophysiology of CF lung disease and hence are not considered targets for CFTR gene transfer.
The lumen of human cartilaginous airways (nasal, tracheal, and bronchial) is normally lined with a pseudostratified mucocili-ary epithelium comprised of ciliated cells and mucus-secreting goblet cells overlying basal epithelial cells (Figure 1). A sub-subpopulation of basal epithelial cells are considered to be the
stem cell precursors for the ciliated and mucus cell phenotypes (20). The morphology of the human lower airway epithelium is distinctly different from the upper airway epithelium, with the lumen of the bronchiolar regions normally lined with simple cubiodal ciliated epithelium with few mucus-secreting cells and no basal cells. Because ciliated epithelial cells are terminally differentiated and basal epithelial cells are absent from the bron-chiolar regions, it has been speculated that mucus-producing cells (Clara cells) may be the progenitors of ciliated cells in these regions. Overall, it is considered that the ciliated cells in both upper and lower airway regions are the target cell types for CFTR gene delivery.
GENE TRANSFER EFFICIENCY TO AIRWAY EPITHELIUM IS LOW DUE TO CELLULAR BARRIERS TO VIRAL ENTRY
Unlike for isolated airway epithelial cells in vitro, it is now apparent that there are several extracellular barriers to AdV on the human airway lumenal surface that result in inefficiency of gene transferin vivo(Figure 1). They include the mucociliary clearance system, the glycocalyceal barrier, the absence of the adenoviral receptors from the airway lumen, and the slow rate of lumenal endocytosis of airway epithelial cells (21–23).
A difficulty in the field of CF lung gene transfer has been identifying a model that recapitulates the phenotypic and mor-phologic characteristics of the human lung airway epitheliumin vivo. Although mouse models are attractive because they are inexpensive and can be easily manipulated genetically, the air-way epithelial cell type distribution in the murine airair-way (with the exception of the murine nasal epithelium) is not reflective of the human. A significant advance in the development of airway models was the generation ofin vitrocell culture models derived from human primary airway epithelial cells obtained from pa-tients with or without CF undergoing lung transplantation (22). Isolated airway epithelial cells grown over a period of 1 to 2 months at an air–liquid interface result in the generation of a pseudostratified, mucociliary airway epithelium that displays similar morphologic and phenotypic characteristics of the in vivohuman cartilaginous airway epithelium (Figure 1). Recent studies have revealed that this model system recapitulates the phenotypic differences that occur between CF and non-CF air-way epithelium: cultures from patients with CF display reduced chloride ion transport, hyperabsorption of sodium ions, the fail-ure to regulate the depth of airway surface liquid, and the dehy-dration of secreted mucus that results in cilial dysfunction and mucostasis (4). Human airway epithelial cell (HAE) culture models have now been used by a number of different groups to test the usefulness of viral and nonviral vectors for delivering genes to the airway epithelium and have been predictive for gene transfer to human airwaysin vivo(22, 23).
Figure 1. Cell type composition of the airway surface epithelium and potential barriers to gene transfer vectors. (A) The airway epithelium is composed of ciliated cells (C), mucus-containing cells (M) overlying basal epithelial cells (B). Gene transfer vectors delivered via the intralu-menal route encounter several barriers that may limit gene delivery to target cells in the airway epithelium: (1) secreted mucus; (2) structure and motion of the cilia; (3) an airway surface fluid; (4) a robust and structurally diverse glycocalyx; (5) an apical membrane that has a lim-ited endocytotic capacity; and (6) epithelial tight junctions that restrict access to basolateral membranes. (B) Histological cross-section of a human airway epithelial cell culture grown at the air–liquid interface showing ciliated, mucus and basal cell composition. Thisin vitromodel of the human airway epithelium recapitulates the physical and biological barriers to gene transfer vectors as seen for the airway epitheliumin vivo.Original magnification:⫻100 (Richardson’s Stain).
expressed afterin vivodosing in less than 20% of the surface epithelial cells; and (2) that basal epithelial cells were efficiently transduced by AdV (17, 24). Subsequently, it was determined that although both columnar and basal epithelial cells express the receptors required for AdV entry into cells, the human cox-sackie B and adenovirus 2 and 5 receptor (hCAR, ) and, ␣v␤3/5integrins (29), expression in columnar cells is restricted to the basolateral domain of the cell membranes (22, 23). In fact, hCAR has been localized to regions associated with epithelial cell tight junctions (30), a location that may restrict access of AdV to hCAR delivered by apical and/or basolateral routes. This observation may explain why in our hands basolateral inocu-lation of HAE by AdV failed to efficiently transduce columnar cell types (Figure 2).
THE GLYCOCALYX AS A RESTRICTION BARRIER FOR ADV DELIVERY TO THE AIRWAY EPITHELIUM
Because the receptors for AdV cellular entry are localized to the basolateral compartments of the epithelium and retargeting
Figure 2. Polarized susceptibility of HAE cultures to AdV infection. Con-focal XZ optical sections of HAE cultures inoculated with AdVGFP and 48 hours later, GFP (green) expression assessed: (A) Inoculation of the apical surface of HAE results in low gene transfer efficiency. (B) Inocula-tion of the basolateral surface results only in efficient gene transfer to basal epithelial cells. (C) Efficient gene transfer to columnar cells after AdV inoculation of the apical surface immediately after tight junctional disruption by sodium caprate. Cilia at the apical surface are identified with anti–␤-tubulin conjugated to Texas Red (red).Original magnification: ⫻63. Figure reproduced with permission from theJournal of Virology(26).
Figure 3. Restriction of AdV access to the apical membrane of airway epithelium by the highly glycosylated and abundant glycocalyx. Sche-matic of the highly glycosylated molecules on the human airway lumenal surface that comprise the glycocalyx layer. Glycoconjugates such as tethered mucins (MUC1, MUC4, MUC16) and proteoglycans (PGs) cover the microvilli-rich airway surface and may attach to and restrict the access of AdV to the apical membrane. The fate of AdV attached to these structures may be engulfment by airway macrophages (M⌽) and/or incorporation into the secreted mucus after shedding of the glycoconjugate-AdV complex from the apical surface.
strategies are dependent of identification of suitable surrogate receptors, we investigated whether redistribution of hCAR to the apical surface of polarized cells would improve the efficiency of AdV-mediated gene transfer. To achieve this, we expressed the external domain of hCAR (containing the AdV attachment site) at the apical surface of polarized epithelia by incorporation
of a glycosylphosphatidylinositol-linker (gpi-hCAR), and found that although this chimeric receptor was expressed at the apical membrane, gene transfer efficiency with AdV still remained low. These studies ultimately identified the lumenal surface glyco-calyx as another barrier to AdV that functions as a “barbed-wire fence” to protect the apical surface including apical receptors from lumenal insult (21). We have now confirmed this barrier function of the airway glycocalyx to AdV with HAE cultures and mouse tracheal epithelium in vivo expressing gpi-hCAR (31). However, the extent of the barrier effect of human airway glycocalyx in restricting AdV access remains controversial (32). The glycocalyx on the airway epithelium lumenal surface is composed of several families of carbohydrate-rich molecules, including glycoproteins (most notably the mucins), proteogly-cans, and glycolipids. A major component of the airway glyco-calyx are the “tethered” mucins, particularly the large (⬎1 mega-dalton), heavily glycosylated MUC1 and MUC4 glycoproteins (33, 34). With respect to airway gene transfer, sialoglycoconju-gates (including MUC1) expressed on the apical surface of polar-ized epithelial cells inhibit AdV-mediated gene transfer (35). Several mucin species will also be present in the mucus in the airway lumen and may act as false attachment sites for AdV thus effectively reducing the amount of AdV that ultimately reaches the apical surface receptors required for viral penetra-tion (Figure 3).
The fate of AdV bound to airway glycocalyx components is speculative. One possibility is that mucins may passively or ac-tively “present” the bound AdV for recognition and engulfment by resident airway macrophages. It has been shown in the murine airway that early after AdV administration to the lung, macro-phages migrate to and internalize AdV that are present in the airway lumen possibly attached to glycocalyx structures (36). It has been estimated that 70 to 90% of AdV are sequestered by airway macrophages 24 hours after intraluminal administration to the lung (37). We predict that glycocalyx components present AdV to incoming macrophages for phagocytosis and degrada-tion. In addition, tethered mucins are shed from the airway surface and incorporate into the soluble mucin layers (34). This property of airway mucins raises the possibility that AdV attached to shed tethered mucins may be eliminated from airways by incorporation into the mucociliary transport system.
OVERCOMING THE LIMITATIONS OF LOW EFFICIENCY OF AdV-MEDIATED GENE TRANSFER
To date, two main strategies to improve the efficiency of gene transfer of AdV after intralumenal delivery have been attempted. One approach is to retarget AdV to nonviral receptors present on the apical surface of lumenal epithelial cells. The other strat-egy attempts to access the basolateral surfaces of the epithelial cells either by nonlumenal delivery of vector or by disruption of epithelial “tight” junctions.
Retargeting AdV to Apical Receptors to Increase Gene Transfer Efficiency
Retargeted AdV can successfully transduce cell types that are usually refractory to AdV infection due to lack of attachment/ entry receptors. The epidermal growth factor receptor, stem cell factor receptor, fibroblast growth factor receptor,␣Vintegrins, and T cell receptors (CD 3) have all been used as surrogate receptors for AdV entry in a variety of cell types (38–40).
Retargeting AdV to receptors on the airway lumen required identification of suitable receptors on the target ciliated cells that are conducive for AdV attachment and entry. Examples of such receptors are members of the 7-transmembrane spanning G protein–coupled receptor family (i.e., P2Y2 purinoceptors,
B2-kinin receptors, and adenosine type 2b receptors). These receptors were identified as putative utile target receptors for redirecting AdV tropism to the surface epithelium of the lung because they are highly expressed on the luminal surface and are internalized into cells upon activation (41). Other receptors proposed as targets for redirecting gene transfer vectors are the urokinase plasminogen activator receptor and the SEC-2 receptor (42, 43).
Retargeting of AdV has so far been achieved by chemically, immunologically, or genetically modifying the AdV capsid coat by incorporating new receptor ligands that can target candidate receptors. As a “proof of concept” study, an hemaggluttin (HA)-epitope tagged P2Y2 receptor expressed at the apical surface of HAE was targeted with bi-specific antibodies consisting of antibodies to AdV fiber-knob protein/HA-tag and in combina-tion with glycocalyx abrogacombina-tion shown to facilitate AdV entry into columnar cell-types (44). Similar retargeting has been shown with chemical conjugation of AdV to receptors via a biotin– streptavidin “bridge” (41).
Retargeted AdV with fiber-knob protein modified to express novel ligands that can interact with target receptors have been developed, and the feasibility of this approach has now been reported by a number of groups (39, 45).
Strategies to Target Endogenous AdV Receptors
The localization of viral uptake pathways to the basolateral surfaces of airway epithelial cells suggests that delivery to this surface could be beneficial for improving gene transfer. Such an approach may also allow targeting of basal epithelial cell subpopulations that may function as stem cells for the columnar epithelium resulting in gene transfer to the lung for the lifetime of the individual. Targeting stem cells is an important consider-ation for gene transfer to the airway epithelium because ciliated cells have a relatively short lifetime (40–90 days), suggesting that strategies to directly target ciliated cells will require read-ministration of vectors every 1 to 3 months.
Access to basal cells/basolateral surfaces after intravenous administration of vectors requires vector dissemination through the lung blood vessel wall and the surrounding connective tissue, as well as penetration of the basal lamina underlying the respira-tory epithelium. Sufficient access to airway epithelial cells via this route has not yet been demonstrated (46).
An alternative approach to target endogenous AdV receptors on columnar cells is transient disruption of epithelial tight junc-tions, thus exposing hCAR, before inoculation by AdV. Walters and coworkers have shown that treatment of the apical surface of HAE with the calcium chelator EGTA or hypotonic solutions (e.g., water) allows for improved AdV-mediated gene transfer (47). The short-chain fatty acid, sodium caprate, has also been shown to increase AdV-mediated gene transfer to HAE and mouse tracheal epitheliumin vivo(27, 48). The ability for tight junctional modulation to potentially allow for AdV access to basal epithelial stem cells remains to be investigated.
However, even with suitable retargeting strategies for AdV the requirement of a vector to circumvent the glycocalyx barrier remains. It is unlikely that effective and safe strategies to abro-gate the airway glycocalyx sufficient for gene delivery improve-ment can be developed. Whether further modifications to the AdV capsid that result in less affinity of the virus for glycocalyx structures could be achieved has not yet been investigated.
IDENTIFICATION OF OTHER VECTORS THAT CIRCUMVENT THE PHYSICAL BARRIERS TO GENE DELIVERY
vectors have been suggested as candidates for CF lung gene transfer vectors and have been rigorously investigated. Adeno-associated viruses, retroviruses, lentiviruses, and liposomal vec-tors have all shown promise in preclinical studies in the lung, and some have been tested in clinical trials. Although most of these vectors can be delivered to the human airway epithelium relatively safely, the efficiency of gene transfer achieved by these vectors has been low, suggesting that barriers to viral entry into the target cell types may exist. Strategies to improve gene transfer efficiency for these other vectors have paralleled the strategies for improving AdV gene transfer. Whether efficiency can be improved sufficiently to shows efficacy of gene transfer in the lungs of patients with CF remains to be determined.
Novel vector types are now being assessed for airway gene transfer that may be more efficient at breaching airway epithelial cell barriers. For example, Sendai virus (SeV), the murine equiv-alent of human parainfluenza virus type 1, has been shown to infect ciliated and nonciliated cells of rodent airways after intra-luminal delivery (49). Human coronavirus 229E has been found to infect human polarized airway epithelia from the apical sur-face (50), although less than 10% of the cells were infected and the large size of this viral genome complicates its use as a vector. HIV-based lentiviral vectors pseudotyped with Ebola virus enve-lope proteins have been shown to efficiently transduce airway epithelial cellsin vitroand murine airwaysin vivo(51), suggesting that combining the efficiency of Ebola virus entry with the poten-tial longer duration of lentivirus-mediated gene expression may provide a useful vector for lung gene transfer strategies. In a similar study, a simian lentiviral vector was pseudotyped with Sendai virus envelope proteins F and HN, which transduced rat polarized tracheal epithelial cellsin vitrofrom the apical surface, albeit with low efficiency (52).
We have recently focused on the paramxyovirus (PV) family of human respiratory viruses. These viruses are a family of enve-loped viruses with nonsegmented negative strand RNA ge-nomes. Human respiratory syncytial virus (RSV) and human parainfluenza virus (PIV) are, respectively, the first and second leading causes of viral respiratory disease in infants and children requiring hospitalization (53). Although most infections are re-stricted to the upper airways and resolve within 1 to 2 weeks without treatment, the entire respiratory tract can be infected, resulting in bronchitis, bronchiolitis, and/or bronchopneumonia, especially in immunocompromised patients. Almost everyone has been infected by PV by 2 years of age, but the immunity induced is typically incomplete and reinfection by the same virus is common although subsequent infections are partially re-stricted and the disease severity reduced (53). With regard to potential gene transfer vectors, it is advantageous that immunity against PV is incomplete and wanes with time.
The tropism of PV for human respiratory epithelium led us to test whether RSV efficiently infected HAE and whether any cell type–specific targeting occurred. The reverse genetics has enabled the rescue of fully recombinant RSV thus allowing for viral genome manipulation and insertion of marker transgenes (54). HAE inoculated with recombinant RSV expressing GFP showed that RSV efficiently infected columnar epithelial cells from the apical surface and exclusively infected ciliated epithelial cells (26). The ability of RSV to transfer genes to the ciliated cells of the airway epithelium after lumenal delivery suggests that this virus may provide a new vector system suitable for disorders of the lung epithelium such as CF lung disease.
RSV infection of respiratory epithelium is however cytotoxic and pulmonary disease due to RSV infection is caused by both direct virus-mediated events and the effects of host immune responses. The most commonly described cytotoxic effect of RSV infection of non-polarized cellsin vitrois giant cell
(syncy-tium) formation, leading to cell death. This effect is likely due to cell membrane expression of RSV glycoproteins that are fuso-genic and can interact with neighboring cells to induce cell–cell fusion (55). However, syncytia formation in respiratory epithe-lium is rarely encountered in pathological specimens from human fatal RSV infection, unless individuals are profoundly immuno-suppressed (53). Indeed our experiments with RSV infection of HAE did not result in ciliated cell syncytia formation, an effect likely due trafficking of the fusogenic viral glycoproteins exclu-sively to the apical surface of ciliated cells where interactions with neighboring cells would be limited. This conclusion was supported by the observation that RSV assembled at and budded exclusively from the apical surface of infected ciliated cells and spread of virus was propagated by the directional flow of airway surface liquid dictated by cilia beat direction (26) (Figure 4). The apical shedding of RSV was not accompanied by gross cytotoxicity and suggests that RSV spread remains within the environment of the lung an observation that may explain why RSV viremia is exceptionally rare. Nevertheless, the cytotoxicity of replication-competent RSV vectors currently limits their use-fulness as gene transfer vectors for “proof of concept” studies. Further study will be required to generate versions of these vectors that could be used to deliver CFTR to the airway epithe-lium in vivo. A wide variety of attenuated viruses have been created and characterized in the development of live attenuated vaccines against PV (56). Attenuation is defined as a reduced ability of the virus to cause disease and most commonly reflects a reduced replication capacity. Replication-attenuated PV vectors would be desirable for a gene transfer vector. The fact that PV vaccine candidates can be safety administered to infants and young children suggests that attenuated PV may be useful for constructing safer gene transfer vectors (57).
Alternatively, pseudotyped vectors generated using glycopro-teins from PV (e.g., RSV pseudotyped lentiviral vectors) could provide vectors that target ciliated cells and have the capacity to improve duration of transgene expression. However, any strat-egy to target ciliated cells with CFTR can only “correct” a CF epithelium for the lifetime of the CFTR-expressing ciliated cells (40–90 days). Therefore, all vectors directed to ciliated cells, even integrating vectors, will ultimately require repeat adminis-tration. In the absence of an identified stem cell for ciliated cells and a suitable vector that can target such a stem cell, the repeat dosing of vectors that target ciliated cells that are safe and can be readministered every 2 to 3 months is an acceptable strategy at the present time.
A gene transfer strategy for the treatment of CF lung disease still struggles to prove itself as a realistic goal, not least because the immature science ofin vivogene transfer was referred to at the outset as “gene therapy.” The promise of a gene therapy set unrealistic goals for this approach, for as we have discovered, there are multiple barriers to achieving efficacious and safe deliv-ery of genes to the lung, of which only a few have been discussed here. The field ofin vivogene transfer has been aided by many different aspects of basic biological and medical research efforts and the eventual realization of a gene therapy for CF lung disease will only take time and a continuation of these efforts.
Conflict of Interest Statement: R.J.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
1. Trapnell BC, Shanley TP. Innate immune responses to in vivo adenovirus infection. In: Curiel DT, Douglas JT, editors. Adenoviral vectors for gene therapy. Boca Raton, FL: Academic Press; 2002. pp. 349–373. 2. Weiss DJ. Delivery of DNA to lung airway epithelium.Methods Mol
3. Riordan JR, Rommens JM, Kerem B, Alon N, Rozmahel R, Grzelczak Z, Zielenski J, Lok S, Plavsic N, Chou JL,et al.Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA.Science1989;245:1066–1073.
4. Matsui H, Grubb BR, Tarran R, Randell SH, Gatzy JT, Davis CW, Boucher RC. Evidence for periciliary liquid layer depletion, not abnor-mal ion composition, in the pathogenesis of cystic fibrosis airways disease.Cell1998;95:1005–1015.
5. Boat T, Welsh MJ, Beaudet AL. Cystic fibrosis. In: Scriver ER, Beaudet AL, Sly WS, Valle D, editors. The metabolic basis of inherited disease. New York: McGraw-Hill; 1989. pp. 2649–2680.
6. Cotton CU, Stutts MJ, Knowles MR, Gatzy JT, Boucher RC. Abnormal apical cell membrane in cystic fibrosis respiratory epithelium. An in vitro electrophysiologic analysis.J Clin Invest1987;79:80–85. 7. Boucher RC. Status of gene therapy for cystic fibrosis lung disease.J
8. Kreda SM, Mengos A, Jensen T, Riordan J, Boucher RC. CFTR expres-sion in normal and Delta-F508 CF airway epithelia.Pediatr Pulmonol
9. Engelhardt JF, Yankaskas JR, Ernst SA, Yang Y, Marino CR, Boucher RC, Cohn JA, Wilson JM. Submucosal glands are the predominant site of CFTR expression in the human bronchus.Nat Genet1992;2:240–248. 10. Drumm ML, Pope HA, Cliff WH, Rommens JM, Marvin SA, Tsui LC, Collins FS, Frizzell RA, Wilson JM. Correction of the cystic fibrosis defect in vitro by retrovirus-mediated gene transfer.Cell1990;62:1227– 1233.
11. Rosenfeld MA, Yoshimura K, Trapnell BC, Yoneyama K, Rosenthal ER, Dalemans W, Fukayama M, Bargon J, Stier LE, Stratford-Perricaudet L,et al.In vivo transfer of the human cystic fibrosis trans-membrane conductance regulator gene to the airway epithelium.Cell
12. Zabner J, Couture LA, Gregory RJ, Graham SM, Smith AE, Welsh MJ. Adenovirus-mediated gene transfer transiently corrects the chloride transport defect in nasal epithelia of patients with cystic fibrosis.Cell
13. Johnson LG, Olsen JC, Sarkadi B, Moore KL, Swanstrom R, Boucher RC. Efficiency of gene transfer for restoration of normal airway epithe-lial function in cystic fibrosis.Nat Genet1992;2:21–25.
14. Caplen NJ, Kinrade E, Sorgi F, Gao X, Gruenert D, Geddes D, Coutelle C, Huang L, Alton EW, Williamson R. In vitro liposome-mediated DNA transfection of epithelial cell lines using the cationic liposome DC-Chol/DOPE.Gene Ther1995;2:603–613.
15. Crystal RG, McElvaney NG, Rosenfeld MA, Chu CS, Mastrangeli A, Hay JG, Brody SL, Jaffe HA, Eissa NT, Danel C. Administration of an adenovirus containing the human CFTR cDNA to the respiratory tract of individuals with cystic fibrosis.Nat Genet1994;8:42–51. 16. Knowles MR, Hohneker KW, Zhou Z, Olsen JC, Noah TL, Hu PC,
Leigh MW, Engelhardt JF, Edwards LJ, Jones KR,et al.A controlled study of adenoviral-vector-mediated gene transfer in the nasal epithe-lium of patients with cystic fibrosis.N Engl J Med1995;333:823–831. 17. Pickles RJ, Barker PM, Ye H, Boucher RC. Efficient adenovirus-medi-ated gene transfer to basal but not columnar cells of cartilaginous airway epithelia.Hum Gene Ther1996;7:921–931.
18. Harkema J, Mariassy A, St George J, Hyde D, Plopper CG. Epithelial cells of the conducting airways: a species comparison. In: The airway epithelium, vol. 55. New York: Marcel Dekker; 1994. p. 3–39. 19. Jeffery P. Form and function of airway epithelium. In: Epithelia: advances
in cell physiology and cell culture. London: Kluwer Academic Publish-ers; 1990. pp. 195–220.
20. Randell SH. Progenitor-progeny relationships in airway epithelium.Chest
21. Pickles R, Fahrner J, Petrella J, Boucher R, Bergelson J. Retargeting the coxsackievirus and adenovirus receptor to the apical surface of polarised epithelial cells reveals the glycocalyx as a barrier to adeno-virus mediated gene transfer.J Virol2000;74:6050–6057.
22. Pickles RJ, McCarty D, Matsui H, Hart PJ, Randell SH, Boucher RC. Limited entry of adenovirus vectors into well-differentiated airway epithelium is responsible for inefficient gene transfer.J Virol1998;72: 6014–6023.
23. Zabner J, Freimuth P, Puga A, Fabrega A, Welsh MJ. Lack of high
affinity fiber receptor activity explains the resistance of ciliated airway epithelia to adenovirus infection.J Clin Invest1997;100:1144–1149. 24. Grubb BR, Pickles RJ, Ye H, Yankaskas JR, Vick RN, Engelhardt JF,
Wilson JM, Johnson LG, Boucher RC. Inefficient gene transfer by adenovirus vector to cystic fibrosis airway epithelia of mice and hu-mans.Nature1994;371:802–806.
25. Johnson LG, Pickles RJ, Boyles SE, Morris JC, Ye H, Zhou Z, Olsen JC, Boucher RC. In vitro assessment of variables affecting the effi-ciency and efficacy of adenovirus-mediated gene transfer to cystic fibrosis airway epithelia.Hum Gene Ther1996;7:51–59.
26. Zhang L, Peeples M, Boucher R, Collins P, Pickles RJ. Respiratory syncytial virus infection of human airway epithelial cells is polarized, specific to ciliated cells, and without obvious cytotoxicity.J Virol2002; 76:5654–5666.
27. Coyne CB, Kelly MM, Boucher RC, Johnson LG. Enhanced epithelial gene transfer by modulation of tight junctions with sodium caprate.
Am J Respir Cell Mol Biol2000;23:602–609.
28. Bergelson JM, Cunningham JA, Droguett G, Kurt-Jones EA, Krithivas A, Hong JS, Horwitz MS, Crowell RL, Finberg RW. Isolation of a common receptor for Coxsackie B viruses and adenoviruses 2 and 5.
29. Wickham TJ, Mathias P, Cheresh DA, Nemerow GR. Integrins␣v␤3 and␣v␤5 promote adenovirus internalization but not virus attachment.
30. Cohen CJ, Shieh JT, Pickles RJ, Okegawa T, Hsieh JT, Bergelson JM. The coxsackievirus and adenovirus receptor is a transmembrane com-ponent of the tight junction.Proc Natl Acad Sci USA2001;98:15191– 15196.
31. Pickles R, Boucher RC. Overcoming the airway epithelial cell glycocalyx-eal barrier to gene transfer.Pediatr Pulmonol Suppl2000;20:152–153. 32. Walters RW, van’t Hof W, Yi SM, Schroth MK, Zabner J, Crystal RG, Welsh MJ. Apical localization of the coxsackie-adenovirus receptor by glycosyl-phosphatidylinositol modification is sufficient for adenovirus-mediated gene transfer through the apical surface of human airway epithelia.J Virol2001;75:7703–7711.
33. Bernacki SH, Nelson AL, Abdullah L, Sheehan JK, Harris A, Davis CW, Randell SH. Mucin gene expression during differentiation of human airway epitheliain vitro. Am J Respir Cell Mol Biol1999;20: 595–604.
34. McNeer RR, Huang D, Fregien NL, Carraway KL. Sialomucin complex in the rat respiratory tract: a model for its role in epithelial protection.
35. Arcasoy SM, Latoche J, Gondor M, Watkins SC, Henderson RA, Hughey R, Finn OJ, Pilewski JM. MUC1 and other sialoglycoconjugates inhibit adenovirus-mediated gene transfer to epithelial cells.Am J Respir Cell Mol Biol1997;17:422–435.
36. Zsengeller Z, Otake K, Hossain SA, Berclaz PY, Trapnell BC. Internal-ization of adenovirus by alveolar macrophages initiates early pro-inflammatory signaling during acute respiratory tract infection.J Virol
37. Worgall S, Leopold PL, Wolff G, Ferris B, Van Roijen N, Crystal RG. Role of alveolar macrophages in rapid elimination of adenovirus vec-tors administered to the epithelial surface of the respiratory tract.
Hum Gene Ther1997;8:1675–1684.
38. Watkins SJ, Mesyanzhinov VV, Kurochkina LP, Hawkins RE. The “ad-enobody” approach to viral targeting: specific and enhanced adenoviral gene delivery.Gene Ther1997;4:1004–1012.
39. Wickham TJ, Carrion ME, Kovesdi I. Targeting of adenovirus penton base to new receptors through replacement of its RGD motif with other receptor-specific peptide motifs.Gene Ther1995;2:750–756. 40. Wickham TJ, Lee GM, Titus JA, Sconocchia G, Bakacs T, Kovesdi I,
Segal DM. Targeted adenovirus-mediated gene delivery to T cells via CD3.J Virol1997;71:7663–7669.
41. Kreda SM, Pickles RJ, Lazarowski ER, Boucher RC. G-protein-coupled receptors as targets for gene transfer vectors using natural small-molecule ligands.Nat Biotechnol2000;18:635–640.
42. Drapkin PT, O’Riordan CR, Yi SM, Chiorini JA, Cardella J, Zabner J, Welsh MJ. Targeting the urokinase plasminogen activator receptor enhances gene transfer to human airway epithelia.J Clin Invest2000; 105:589–596.
43. Ziady A, Keefer R, Ferkol T, Davies P. Serpin enzyme complex receptor targeted DNA complexes deliver genes to airway epithelia. (abstract)
44. Pickles R, Kreda S, Olsen J, Johnson L, Gerard R, Segal D, Boucher R. High efficiency gene transfer to polarised epithelial cells by retargeting adenoviral vectors to P2Y2 purinoceptors with bispecific antibodies.
45. Krasnykh VN, Mikheeva GV, Douglas JT, Curiel DT. Generation of recombinant adenovirus vectors with modified fibers for altering viral tropism.J Virol1996;70:6839–6846.
46. Lemarchand P, Jones M, Danel C, Yamada I, Mastrangeli A, Crystal RG. In vivo adenovirus-mediated gene transfer to lungs via pulmonary artery.J Appl Physiol1994;76:2840–2845.
47. Walters RW, Grunst T, Bergelson JM, Finberg RW, Welsh MJ, Zabner J. Basolateral localization of fiber receptors limits adenovirus infection from the apical surface of airway epithelia.J Biol Chem 1999;274: 10219–10226.
48. Johnson LG, Vanhook MK, Coyne CB, Haykal-Coates N, Gavett SH. Safety and efficiency of modulating paracellular permeability to en-hance airway epithelial gene transfer in vivo.Hum Gene Ther2003;14: 729–747.
49. Yonemitsu Y, Kitson C, Ferrari S, Griesenbach U, Farley R, Steel R, Scheid P, Judd D, Kato A, Hasan M,et al.Recombinant sendai virus can efficienctly transfect airway epithelium in vivo.Mol Ther2000;1: A697.
50. Wang G, Deering C, Macke M, Shao J, Burns R, Blau D, Holmes K, Davidson B, Perlman S, McCray P. Human coronavirus 229E infects polarized airway epithelia from the apical surface.J Virol2000;74: 9234–9239.
51. Kobinger GP, Weiner DJ, Yu QC, Wilson JM. Filovirus-pseudotyped lentiviral vector can efficiently and stably transduce airway epithelia in vivo.Nat Biotechnol2001;19:225–230.
52. Kobayashi M, Iida A, Ueda Y, Hasegawa M. Pseudotyped lentivirus vectors derived from simian immunodeficiency virus SIVagm with envelope glycoproteins from paramyxovirus. J Virol 2003;77:2607– 2614.
53. Collins PL, Chanock RM, Murphy BR. Respiratory syncytial virus. In: Knipe DM, Howley PM, editors. Virology, 4th ed. Philadelphia: Lip-pincott-Raven Publishers; 2001. pp. 1443–1485.
54. Collins PL, Mink MA, Stec DS. Rescue of synthetic analogs of respiratory syncytial virus genomic RNA and effect of truncations and mutations on the expression of a foreign reporter gene.Proc Natl Acad Sci USA
55. Roberts SR, Compans RW, Wertz GW. Respiratory syncytial virus ma-tures at the apical surfaces of polarized epithelial cells.J Virol1995; 69:2667–2673.
56. Karron RA, Buonagurio DA, Georgiu AF, Whitehead SS, Adamus JE, Clements-Mann ML, Harris DO, Randolph VB, Udem SA, Murphy BR,et al.Respiratory syncytial virus (RSV) SH and G proteins are not essential for viral replication in vitro: clinical evaluation and molecular characterization of a cold-passaged, attenuated RSV subgroup B mu-tant.Proc Natl Acad Sci USA1997;94:13961–13966.